Diachronic vs. synchronic universals/tendencies

[manaster at umich.edu]

----------------------------Original message----------------------------
In the discussion of this topic, which started with
Scott Delancey's defense of the idea that many or
even all (?) synchronic universals or tendencies
have a diachronic explanation,
I think there are two separate points involved
which we have all been confusing (myself included).
 
One,
as suggested by Richard Janda, there is a case to
be made (I am being deliberately conservative) that
all linguistic universals or tendencies ultimately
depend on our biological endowment and hence are
"synchronic".  But as far as I can see this does not
detract from the force of Scott Delancey's observation,
since it is logically possible that the way these
innate mechanisms manifest themselves is precisely
in the kind of language effect that Scott was talking
about.  So the question of whether certain word order
tendencies are the way they are because of certain
proposed diachronic tendencies is to my mind quite
separate from the deeper question of how the
diachronic tendencies arise from our wiring.
(Of course, I must add that I do not believe that
we are in position in linguistics
to really pinpoint what the wiring is, unlike in
research on bird songs, etc., and so we do NOT
in fact know whether things the way they are
because of a particular bit of putative wiring,
in general.  The fact that something appears to
occur univeresally in all languages known to
the author of any given article, whether that
is just English or a hundred languages, does
not in any way shape or form mean that it
is hardwired into human beings.)
 
Now, given that the innatist (everything is
synchronic) position is (a) possibly right
but this could is nonetheless consistent with
everything Scott says and (b) is not
either verifiable or falsifiable given
the limitations on what kind of research
linguists can do (e.g., we are not allowed
to deafen babies at birth), I think I would
like to get back to the other, more modest
question of whether Scott's model actually
fits the facts--regardless of whether it
ultimately reduces to some sort of innate
synchronic basis.  As I see, two kinds of
things are relevant here.
 
One, consider
Mandarin Chinese, where there seems to be
a clear trend towards SOV because it
often now prefers S ba O V to simple
SVO (where ba is some kind of verb),
but as ba O illustrates new kinds of
adpositional phrases made with a verb
that becomes an adposition are in fact
pre- and not postpositional, so here
we have a diachronic process which seems
designed to yield an SOV language with
prepositions of verbal origin, the opposite
of what Scott suggests.  And if this is
possible and if in addition tehre is no
purely synchronic tendency to make OV
languages postpositional, then we would
have expect no correlation between OV
and postpositions. But we do.
 
Two, the other
kind of example that is crucial is what
I referred to in an earlier message, the
tendency of languages to "fix" synchronically
unnatural systems.  I dont at the moment
have the data at my disposal (or else I am
suffering a lapse of memory) needed to
demonstrate that this happens in word
order phenomena.  But what of my other
examples, such as the ones in phonology
or the fact that the disappearance of
active-voice forms in certain tenses and
moods in Indo-Iranian languages does not
lead to the "unnatural" situation that
you have to use passives only but rather
to the passives are immediately reinterpreted
as actives and thus you have surface
ergativity.  You seem to have a diachronic
process, which itself is apparently quite
natural, which favors the loss of active
voice forms in certain tenses/moods (and
this process may indeed have some
unviersal basis since it always seems to
be the same tenses/moods that are involved),
but it produces an unnatural situation and
the language is forced to change FURTHER.
In phonology, of course, we seem to have
lots of examples of this: somehow all
IE languages in which the glottalized
stops shift to voiced manage to get lots
of [b]'s from somewhere even though
there are almost no glottalized *[p']'s
in PIE.  So something forces a language
to fill that gap (or near-gap).  Consider
another instance: English has lost in
most of its dialects the distinction
between sg. and pl. in the 2nd person,
but the result is that it is practically
obligatory in many dialects to say SOMETHING
in addition to 'you' to make the plural, e.g.,
you folks, you guys, the two of you, all of you,
you kids, you ... all, etc.   So again it
seems as though there is a synchronic
force which tries to make us not merely
use you (perceived as primarily sg.) when
we intend the plural.
 
I am not at all sure that this is right,
and I intend this purely as points for
discussion.  Maybe there IS a way to
account for all this Scott's way, but
I do not as yet see that.
 
AMR